Practical Applications and Approaches in GeoAI

GEOG 288KC: Geospatial Foundation Models and Applications

Fall 2025

Practical skills for geospatial AI applications

Course Description

This accelerated, hands-on seminar provides practical skills for working with state-of-the-art geospatial foundation models. Students learn to access, process, and analyze satellite imagery using modern tools, apply foundation models to real-world problems, and implement independent projects in environmental monitoring and analysis.

By the end of the course, students will be able to:

• Access and process satellite imagery using STAC APIs and cloud platforms
• Apply preprocessing pipelines for multi-temporal remote sensing data
• Train custom CNN models for land cover classification and change detection
• Load and fine-tune pretrained geospatial foundation models
• Implement spatiotemporal analysis for environmental monitoring applications
• Design and execute independent research projects using geospatial AI

Getting Started with the UCSB AI Sandbox

Here are detailed instructions for setting up the class environment on the UCSB AI Sandbox, including foundation model installation and GPU optimization. This should all be taken care of for the class, but could be helpful if you are interested in deploying our class infrastructure on a different server or a local machine.

Course Structure: 10-Week Format

flowchart TD
    subgraph Phase1 ["📚 Phase 1: Structured Learning (Weeks 1-5)"]
        direction LR
        W1["🚀<br/>Week 1<br/>Core Tools<br/>Data Access"] --> W2["⚡<br/>Week 2<br/>Preprocessing<br/>Cloud Processing"]
        W2 --> W3["🤖<br/>Week 3<br/>ML & CNNs<br/>Classification"]
        W3 --> W4["🏗️<br/>Week 4<br/>Foundation Models<br/>Feature Extraction"]
        W4 --> W5["🔧<br/>Week 5<br/>Fine-Tuning<br/>Transfer Learning"]
    end

    subgraph Phase2 ["🎯 Phase 2: Independent Project Development (Weeks 6-10)"]
        direction LR
        W6["📋<br/>Week 6<br/>Project Proposals<br/>Planning"] --> W7["🔬<br/>Week 7<br/>Initial Implementation<br/>MVPs"]
        W7 --> W8["⚙️<br/>Week 8<br/>Development<br/>Refinement"]
        W8 --> W9["📊<br/>Week 9<br/>Analysis<br/>Results"]
        W9 --> W10["🎉<br/>Week 10<br/>Final Presentations<br/>Deliverables"]
    end

    Phase1 --> Phase2

    style Phase1 fill:#e3f2fd
    style Phase2 fill:#fff3e0
    style W1 fill:#bbdefb
    style W3 fill:#bbdefb
    style W5 fill:#bbdefb
    style W6 fill:#ffe0b2
    style W8 fill:#ffe0b2
    style W10 fill:#c8e6c8

📚 Phase 1: Hands-On Practice (Weeks 1-5)

• Week 1: Core Tools and Data Access - STAC APIs, satellite data visualization, NDVI calculation
• Week 2: Remote Sensing Preprocessing - Cloud masking, reprojection, temporal compositing
• Week 3: Machine Learning on Remote Sensing - CNN training, land cover classification, model comparison
• Week 4: Foundation Models in Practice - Loading pretrained models, feature extraction, practical applications
• Week 5: Fine-Tuning & Transfer Learning - Linear probing vs. full fine-tuning, adaptation strategies

🎯 Phase 2: Independent Project Development (Weeks 6-10)

• Week 6: Project Proposals & Planning - Define scope, methodology, and expected outcomes
• Week 7: Initial Implementation - Develop minimum viable product (MVP), early results
• Week 8: Development & Refinement - Expand functionality, optimize performance
• Week 9: Analysis & Results - Generate final results, prepare visualizations
• Week 10: Final Presentations - Present completed projects, peer review, submission of deliverables

Course Sessions

• Weekly sessions: see navbar → 💻 weekly sessions

Teaching Team

Instructor

Kelly Caylor
Email: caylor@ucsb.edu
Learn more: Bren profile

TA

Anna Boser
Email: anaboser@ucsb.edu
Learn more: Bren profile